Murder Mysteries, Mock Interviews, and the GenAI Revolution

by Diana Drake

What happens when generative AI and innovation collide? Tech fire.

At the Wharton Hack-AI-Thon, co-hosted by AI at Wharton, Analytics at Wharton and the Wharton AI and Analytics Club, University of Pennsylvania and Wharton School students worked in small teams to create something new. In this hackathon with a twist, they had a week to engineer either an app or prompt using generative AI, which is the type of artificial intelligence and machine learning that uses models to create new content, such as text and images (think ChatGPT).

Chatbots and Small Talk

For competitors, the Wharton Hack-AI-Thon was a chance to dig into a discipline that is truly on the cutting edge – of education and society.

“GenAI represents a rapidly advancing and exciting area within data science,” observes Xinmei “Cindy” Feng, a data science grad student at Penn who was the leader of the winning “Prompt Engineering” team that was charged with innovating around the prompt engineering of GenAI. Prompts are essential to GenAI because they act as a bridge between what a user wants and what the machine understands. “As a data science student, I’m naturally drawn to this field and saw the event as a fantastic opportunity to gain practical experience in an increasingly vital sector.”

Cindy (in orange) and her teammates at the Hack-AI-Thon.

Hanging out with an enthusiastic group of AI innovators was also pretty inspiring, added Kevin Li, leader of the winning “Technical Track” team, which had to create an app using the OpenAI API (Application Programming Interface). “It was fascinating to see how different teams harnessed AI to tackle diverse challenges,” notes Kevin, a Wharton MBA and one-time Harker School grad with experience in investment banking and startups. “What stood out was the creativity in applying AI to different domains like tax, health care and education, showcasing the versatile impact of AI across sectors.”

Several pioneering projects lit up the digital platform at the end of the Hack-AI-Thon week, including Moodshaker, an app that tailors cocktail suggestions based on the user’s mood, and Small Talk Master, engineering GenAI prompts to help international students with more effective social interactions.

The first-place teams (led by Cindy and Kevin), took home $1,000 each. They look like this:

Detective GPT, an AI chatbot that won in the prompt engineering category, allows users to quickly develop interactive detective games like murder mysteries. The game application offers a high level of customization and interactive freedom, allowing users to select their style of gameplay. Once a style is chosen, GPT technology generates a unique detective game in seconds. Users then interact with the game through the ChatGPT interface, taking on the role of a detective tasked with solving a murder. The game mimics real-world investigative processes, presenting various challenges such as debunking alibis, identifying the murderer, and solving locked-room scenarios.

“The concept for Detective GPT arose from our desire to innovate beyond the well-trodden paths of AI,” says Cindy, adding that her team plans to further develop their project with  more data and advanced large language model techniques. “While AI is already proficient at generating text, images and music rapidly, its application in gaming, particularly in creating dynamic and interactive narratives, is less explored,” she continues. “We identified visual novel games as an ideal genre to benefit from GenAI due to their reliance on compelling scripts and interactive text-based communication…AI technology, particularly GenAI, informs and enhances our game by accelerating content creation – scripts can be developed swiftly using AI, tapping into vast datasets to generate dialogue and plot twists.”

AiMI: Created by the winners of the technical track, AiMI is an advanced interview assistant that helps users improve their interview skills and address common challenges in interview preparation. It designs customized interview questions and scenarios, offers verbal and non-verbal feedback, and accepts input in the form of text or video.

Kevin Li (left, center) and his teammates.

“AiMI is an AI mock interviewer,” says Kevin, who is heading to Japan when he graduates in May 2024 to work in e-commerce as an Amazon product manager. “My four other teammates and I have experience interviewing for software engineering, consulting, tech, and investment banking roles, to name a few industries. When we met up as a team to discuss ideas, one common pain point was preparing for these job interviews. AiMI was conceived to address the challenges many face in preparing for job interviews. We leveraged AI to analyze user responses for content relevance, speech clarity, and engagement, enabling tailored coaching that improves interviewing skills.”

What It Means to Be Human

While both Cindy and Kevin have collaborated with their teams on developing practical GenAI applications, they are taking much grander visions away from the Hack-AI-Thon’s exchange of ideas and peer innovation. “The projects I’ve seen showcase AI’s potential to solve real-world problems and predict a future where technology significantly amplifies human productivity and creativity,” notes Kevin.

For Cindy, it’s clear that AI is not just a tool, but a transformative force that is reshaping our world. “This transformation presents us with profound questions and opportunities,” observes Cindy, who hopes to eventually integrate AI into the financial sector to automate and enhance financial processes and structures. “How do we leverage this technology responsibly? How can we ensure that AI enhances human welfare without compromising our values or autonomy?”

Understanding and adapting to AI technology will be pivotal in navigating the future job market and contributing effectively to society, Cindy suggests. “The rise of AI signifies a future where technological literacy will be crucial. It also underscores the importance of cultivating soft skills that AI cannot replicate, such as interpersonal communication and problem-solving in complex, unstructured situations. This is a time for young innovators to engage with AI not just as users but as shapers of this technology, ensuring it grows in ways that are beneficial and aligned with our highest human ideals. This era is more than a technological revolution; it’s a moment to define what it means to be human in an increasingly automated world.”

Conversation Starters

What are your thoughts on Detective GPT and AiMI? What questions would you like to ask the project teams? Post them in the comment section of this article.

Cindy notes that AI as a transformative force is an opportunity for us to think critically about how we can “ensure that AI enhances human welfare without compromising our values or autonomy.” What are your thoughts on this profound question?

Are you tech literate, especially when it comes to GenAI? Do you feel encouraged or discouraged to use it?

Leave a Reply

Your email address will not be published. Required fields are marked *